Second Biennial Hearing Aid Research and Development Conference
September 22-24, 1997
National Institutes of Health
Evaluation of Audibility Model Using Clinical Patients*
ReSound Corporation, Redwood City, California
Speech hearing loss has been studied by comparing speech recognition for hearing impaired groups and noise-masked normal hearing groups at equal masked thresholds. One conclusion of these studies is that the primary source of speech hearing loss is the loss of audibility. A miss of a test score from the prediction of an audibility model (AI or SII) may be explained in terms of proficiency factor such as experiences in listening the test and skills in decoding the speech message. This paper examines how accurate the audibility model (with a proficiency assumption) is in evaluating speech hearing loss for clinical patients. Speech recognition scores for CID W22 were tested for 236 patients. Both left and right ears were evaluated for each patient. Two scores were obtained for each ear: One at a low level where a score of 50% was predicted based on the SII model and another at a high level where a maximum SII score was predicted. These scores are compared with the predictions using the audibility model for the CID W22 material. It was found that speech scores for 60% of the patients can be explained by the SII model. About 13% can be explained by the SII model with the addition of a proficiency assumption The remaining 27% cannot be explained by the audibility model. More details will be presented and discussed.
[*This work is partially done at Massachusetts Eye and Ear Infirmary, Boston, Massachusetts]
Return to Conference Abstracts List